Tunable Large Pore Mesoporous Carbons for the ... - ACS Publications

Subscriber access provided by CORNELL UNIVERSITY LIBRARY

Article

Tunable Large Pore Mesoporous Carbons for the Enhanced Adsorption of Humic Acid Wannes Libbrecht, a Verberckmoes, Joris W. Thybaut, Pascal Van Der Voort, and Jeriffa De Clercq Langmuir, Just Accepted Manuscript • Publication Date (Web): 15 Jun 2017 Downloaded from http://pubs.acs.org on June 16, 2017

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Langmuir is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 29

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

Tunable Large Pore Mesoporous Carbons for the Enhanced Adsorption of Humic Acid Wannes Libbrechta,b,c, An Verberckmoesa , Joris W. Thybautb, Pascal Van Der Voort c and Jeriffa De Clercqa*

a

Industrial Catalysis and Adsorption Technology (INCAT), Faculty of Engineering

and Architecture, Ghent University, Valentin Vaerwyckweg 1, 9000 Ghent, Belgium b

Laboratory for Chemical Technology, Faculty of Engineering and Architecture,

Ghent University, Technologiepark 914, 9052 Zwijnaarde, Belgium c

Center for Ordered Materials, Organometallics and Catalysis (COMOC), Department

of Inorganic and Physical Chemistry, Ghent University, Krijgslaan 281-S3, 9000 Ghent, Belgium

*Correspondence

to: Jeriffa De Clercq, Industrial Catalysis and Adsorption

Technology (INCAT), Faculty of Engineering & Architecture, Ghent University, Valentin Vaerwyckweg 1, 9000 Ghent, Belgium

E-mail address: [email protected]

Keywords Mesoporous carbons – Soft template – Adsorption – Humic acid

ACS Paragon Plus Environment

Langmuir

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Abstract Tunable large pore soft templated mesoporous carbons (SMC) were obtained via the organic self-assembly of resorcinol/formaldehyde with the triblock copolymer F127 and by investigating the effect of carbon precursor to surfactant (p/s) ratio and carbonization temperature on the material characteristics. The p/s ratio and carbonization temperature were varied respectively from 0.83 to 0.25 and from 400 °C to 1200 °C. The resulting SMCs had various average pore sizes, tunable from 7 up to 50 nm. At lower p/s ratios, the pore size increased, pore size distributions broadened and pore volumes increased. An increase of hydrophobicity was observed at elevated carbonization temperatures. A 2D hexagonal ordered SMC with a narrow pore size distribution was obtained at a ratio of 0.83. Lower ratios (0.5 and 0.25) transformed into disordered porous carbons containing micro-, meso- and even macropores. The SMCs were tested for adsorption of a large organic molecule, humic acid (HA), from water. The material characteristics that significantly affected HA adsorption capacity, were pore size and mass percentage (wt%) carbon. The novel SMCs showed an unprecedented high adsorption of HA in the entire molecular weight distribution range. SMCs with large mesopores resulted in higher maximum adsorption capacities and improved HA adsorption kinetics compared to activated carbon.

1. Introduction Mesoporous carbons have been studied extensively because of their wide variety of applications, ranging from catalysts, gas storage hosts, electrode materials to adsorbents [1–7]. The ability to control material characteristics such as specific surface area, pore volume, pore size and morphology, give rise to this multitude of


Page 2 of 29

Page 3 of 29

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

applications. The predominant synthesis route is via soft template [8]. A good interaction between carbon precursor and surfactant is required to form an ordered mesoporous structure. This soft templated organic-organic self-assembly method is mostly performed by polymerizing phenol or resorcinol with formaldehyde around the micelles of a surfactant, e.g., pluronic F127 or P123. The amphiphilic surfactant creates structured micelles that adopt tailored symmetries and interact with the carbon precursor. Via an acid or base catalyzed mechanism the precursors react into a phenolic Novolac or Resol resin [9]. The mesostructure formation is aided by the evaporation induced self-assembly (EISA) at acid or neutral conditions. This is followed by thermopolymerization to give a highly cross-linked composite. In a final carbonization step the surfactant is removed and the polymer transforms into a stable carbon framework. The pore size control is a challenge within mesoporous carbon synthesis. In soft templated synthesis the choice of surfactant is most crucial to determine the pore size. The most frequently used pore-forming components are the triblock copolymers (F127, P123 and F108) with a polypropylene oxide center block and polyethylene oxide ends of varying sizes. These surfactants almost always result in pore sizes between 4 and 10 nm. A large family of mesoporous carbons with good morphology control was synthesized by Meng et al., using both F127, P123 and F108 [10]. The morphologies were 3D bicontinuous cubic (Ia3d), 2D hexagonal (p6m) and bodycentered cubic (Im3m) respectively FDU-14, -15 and -16. These different morphologies had pore sizes ranging from 3.5 nm to 7.1 nm. For the soft templated synthesis route which uses strong acids, such as HNO3 or HCl, as polymerization catalyst [11–13] together with F127, Górka et al. found that an increased acid concentration shifted the channel-like pores to cage-type pores and increased the pore


Langmuir

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

size up to 12 nm [14]. By adjusting the resorcinol to F127 ratio and polymerization time, Tanemura et al. were able to create mesopores of 12.5 nm [15]. Recently, the adjustment of the resorcinol-formaldehyde resin to F127 ratio and use of a very low water to ethanol ratio gave porous carbons consisting of aggregated particles with an average pore size of 56 nm in between these particles [16]. Another approach to increase the pore size, is the use of high-molecular-weight amphiphilic block copolymers. These non-commercial surfactants create larger micelles as their hydrophobic polymer chain length increases, e.g., poly(ethylene oxide)-block-poly(3-caprolactone) (PEO-b-PCL) surfactants were synthesized with different CL chain lengths. By varying the surfactant and the p/s ratio, mesopores between 6.8 and 15.6 nm were obtained [17]. Deng et al. synthesized FDU-18 with a pore size of 10 nm with poly(ethylene oxide)-b-polymethyl methacrylate (PEO–PMMA) [18], while the use of the larger diblock copolymer poly(ethylene oxide)-block-poly(styrene) (PEOb-PS) increased the pore size of FDU-18 to almost 30 nm [19]. Vorasurf 504, a commercial triblock copolymer with a poly(butylene oxide) center and poly(ethylene oxide) tails (PEO–PBO–PEO) introduced 27 nm pores in an EISA resorcinolformaldehyde system [20]. Large mesoporous carbons find their applications as support for nanoparticles, catalysts and as adsorbents for larger molecules e.g. enzymes, bulky dyes or humic acid (HA) [21–24]. HA adsorption was selected in this study. Humic acid (HA) represents natural macromolecular and dissolved organic matter formed by decomposition of natural organic matter. It comprises of organic molecules (MW of 350-50000 g/mol), which are ubiquitous in surface water and ground water. HA may cause color, taste, odor problems, and be harmful to aquatic organisms, such as algae, fish and invertebrates. Furthermore, the HA in drinking water reacts with Cl2, a


Page 4 of 29

Page 5 of 29

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

popular disinfectant, to form trichloroalkanes. The HA acts as the precursor of these strongly carcinogenic disinfection byproducts (DBPs) generated in chlorinated or chloraminated drinking waters [11,25–27]. . The two main HA adsorption mechanisms are based on electrostatic or hydrophobic interaction. The electrostatic interactions are observed between the negatively charged functional groups of HA, e.g., carboxyl, carbonyl and hydroxyl groups, and the positive charges of, e.g., soil minerals [28] or aminopropyl-functionalized SBAmaterials [29]. SBA-15 itself is not suited for HA adsorption, but the introduction of 1.8 mmol/g aminopropyl functionalities increased the maximum adsorption capacity to 117.6 mg/g, which is considerably better than activated carbon (AC) with 52.4 mg/g [29]. Soil minerals are often used as low-cost adsorbents. The pillared bentonite with a specific surface area and basal spacing of 111.3 m²/g and 1.98 nm exhibited a high HA adsorption capacity of 537 mg/g, however, the adsorption depends on the pH, with the strongest adsorption occurring at a pH of 4 [30]. This is due to the electrostatic interactions between the soil mineral and HA. In carbon sorbents, hydrophobic interactions through π-stacking are observed between the surface and the hydrophobic moieties (aromatic domains) of HA. [23,24,31]. Activated carbons possess the right type of surface to interact with HA, but the small pores often prohibit a sufficient adsorption capacity. The large size of the HA molecules results in their insufficient removal by AC traditionally used in waste water treatment plants [32]. Kołodziej et al. performed a post treatment of activated carbon to introduce mesoporosity around 25 nm [31]. This resulted in an increased maximum adsorption capacity of HA up to 140 mg/g. A hard templated mesoporous carbon CMK-3 with pore size of 3.7 nm was also tested for HA adsorption with a maximum capacity of 137 mg/g [24]. Because of the various MW and size of HA molecules, larger pores


Langmuir

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

presumably increased the maximum adsorption capacity. The mesoporous CMK-3 was able to adsorb the large HA molecules while AC was not able to do so [24]. The same size exclusion effect was observed by Liu et al. who investigated the HA adsorption by covalent triazine frameworks (CTF). By changing the monomer from 1,4-dicyanobenzene to 4,4-biphenyldicarbonitrile, the CTF pore size was increased resulting in CTF-1 and CTFDCBP respectively. The microporous CTF-1 (< 2 nm) showed almost no HA adsorption in contrast to the mesoporous CTFDCBP with pores up to 14 nm which had an adsorption capacity of 140 mg/g [26].

In this study, the effect of carbon precursor to F127 surfactant (p/s) ratio and carbonization temperature on the mesoporous carbon material characteristics and HA adsorption was investigated. By fine-tuning the p/s ratio, mesoporous carbons with a high degree of control in both pore volume as well as pore size can be synthesized. To our knowledge mesoporous carbons with an average pore size of 40 nm and a broad pore size distribution with pores up to 80 nm have not yet been synthesized by using the commercial F127 surfactant.

2. Experimental 2.1. Mesoporous carbon synthesis An acid catalyzed EISA soft templated mesoporous carbon synthesis method was used. 2.2 g resorcinol (≥99%, Sigma Aldrich) and varying amounts of F127 (Sigma Aldrich, 2.65 g - 8.8 g), resulting in p/s mass ratios of 0.83, 0.5 and 0.25, were selected and mixed with 9 mL ethanol (≥ 99.8% purity, VWR) and 9 mL HCl (3M) and stirred for 15 minutes at room temperature in a closed reactor. Then, 2.6 g formaldehyde (Formalin 37 wt% in water) was slowly added and the solution was stirred again for 15 minutes in a closed reactor. The solution was subsequently poured


Page 6 of 29

Page 7 of 29

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

into a large petri-dish to evaporate the ethanol at room temperature for 6 h. Next, the film was scraped from the petri-dish and put into a Nabertherm muffle oven for curing at 60 °C for 12 h. The cured resin was calcined and carbonized under N2-flow in a Thermolite tubular furnace. In a first heating step of 1 °C/min to 350 °C for 2 h, the surfactant was removed. A second heating step of 2 °C/min to a carbonization temperatures of 400 °C, 800 °C or 1200 °C was performed and the temperature was subsequently kept constant for 3 h. The carbon monolith was then crushed and sieved through a 150 mesh filter. Samples were named SMC-x-y, where x and y represent the carbonization temperature and p/s ratio respectively.

2.2. Adsorbent characterization Powder X-ray diffraction (PXRD) patterns of the adsorbents were collected on a Thermo Scientific ARL X’Tra diffractometer, operated at 40 kV, 40 mA using Cu Kα radiation with 0.15418 nm wavelength. Nitrogen gas sorption experiments were conducted at 77 K using a Micromeritics Tristar 3000. Samples were vacuum-dried at 120 °C overnight prior to analysis. Total pore volume Vtot was calculated as the amount of nitrogen adsorbed at a relative pressure of 0.95. Vmicro was calculated from the V–t plot. The total surface area SBET was determined using the Brunauer– Emmett–Teller (BET) method. The average pore size (Dp) and the pore size distribution (PSD) for SMC-x-0.83 were calculated from the desorption branch using the BJH method. Elemental analysis was performed with a CHNS-O analyser Thermo Scientific Flash 2000, with a TCD detector, using the Eager Experience software. SEM-images were taken on an EOL JSM 7600F FEG SEM operating at 10 kV.


Langmuir

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

2.3. Adsorption tests To dissolve HA (99+% purity, Sigma Aldrich) in water, the pH was increased by adding 0.1M NaOH solution to a pH of 12 and the HA solution was placed in an ultrasonic bath for 30min. Next, the pH was decreased to 7 by addition of 0.1M HCl. The HA-concentration in solution after adsorption was determined by filtering the mixture through a 0.45 µm PET syringe filter and analyzing the HA-concentration of the filtrate with a Thermo scientific Evolution 60 UV/VIS spectrophotometer at a wavelength of 254 nm, blank samples without adsorbent were measured for each adsorption run. The adsorption capacity of all synthesized adsorbents was determined by adding 10 mg adsorbent to 50 mL HA-solution with a concentration of 500 mg/L. Samples were placed in a thermostatic shaking device, Infors HT, multitron standard, for 7days at 25 °C and stirred at 200 rpm. The remaining HA-concentration after adsorption experiment was measured to determine the amount retained on the adsorbent. All adsorption experiments were performed in triplicate. To investigate the relationship between the material characteristics and the adsorption capacity, a stepwise regression analysis was performed to select the significant parameters. Forward selection of the significant parameters was performed by comparing the models for the addition of each parameter using the Akaike information criterion [33]. Single, interaction and quadratic effects were investigated. The significance of each effect is tested based on the H0 hypothesis which assumes that a given effect is not significant. Assessment of effect significance is done based on its p-value, the probability for falsely rejecting the H0 hypothesis: an effect is considered to be significant when p < 0.05. Adsorption kinetics were measured by adding 100 mg adsorbent to 500 mL of 40 mg/L HA-solution. Both AC (Desotec, Belgium) and SMC-800-0.25 were used.


Page 8 of 29

Page 9 of 29

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

All used mesoporous carbons and activated carbon were sieved through a 170 mesh sieve, particle smaller than 88 µm were used for all adsorption experiments. The experiments were carried out at room temperature (RT) and the mixture was stirred at 700 rpm. Samples were taken at different time intervals and analyzed. The adsorption capacity at time t (qt) was calculated according to Eq.1 with C0 and Ct the initial HAconcentration and the HA-concentration at time t (mg/L), m the adsorbent mass (g) and V the solution volume (L). The experimental data were fitted with the pseudo second order and the Weber Morris model.

qt =

(C0 − Ct )V m

(1)

Adsorption isotherm experiments were performed by adding 10 mg adsorbent to 100 mL HA-solutions ranging from 5 to 1000 mg/L. Subsequently, these mixtures were placed in a thermostatic shaking device, Infors HT, multitron standard, for 7days at 25 °C and stirred at 200 rpm to reach equilibrium. The adsorption capacity at equilibrium (qe) was calculated according to Eq. 1, with Ct replaced by the equilibrium HA-concentration Ce (mg/L). The experimental data were fitted with the Langmuir and Freundlich model. Molecular size distributions of a 80 mg/L HA-solution before and after adsorption with AC and SMC-800-0.25, were analyzed via liquid chromatography organic carbon detection (LC-OCD). The adsorption conditions were 50 mL HA-solution and 50 mg adsorbent at 25 °C and 200rpm for 7days. The effect of different initial pH values on the adsorption capacity of SMC-800-0.25 was investigated. 10 mg SMC800-0.25 was added to 50 mL of HA-solution (40 mg/L) for 7 days at 25 °C and 200 rpm with initial pH values of 3, 5, 7, 9 and 11. The experiment was performed in triplicate.


Langmuir

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

3.

Results and Discussion

3.1. Effect of p/s ratio on material characteristics Nitrogen sorption measurements (Fig. 1) show a type IV isotherm with H1 hysteresis for the SMC-800-y materials indicating mesoporous materials. The narrow hysteresis for all isotherms indicates that the pores are easily accessible and there is no pore blocking, ink-bottle pores or cage like pores. For AC, a type I isotherm indicates the presence of predominantly micropores. Lowering the p/s ratio shifts the hysteresis to a higher value of p/p0, indicating an increase in pore size as seen in Fig. 1. Together with an increasing average pore size, an extensive broadening of the pore size distribution is observed. This broadening could be attributed to the destabilization of the mesostructure. An increase in pore volume can be explained by a change in the modified packing parameter. This parameter has been used to describe the surfactant organization in the self-assembly arrays and to predict the resultant mesostructures. Lower p/s ratios lead to a large modified packing parameter which creates a 3D pore network [16,34]. At low p/s ratios the low amount of carbon precursor will be insufficient to form 2D hexagonal walls around each micelle. In case of SMC-8000.25 this creates a broad-range mesoporous material with additional macropores up to 80 nm. The average pore size for SMC-800-0.83, -0.5 and -0.25 were respectively 6.2, 23.7 and 43.8 nm as shown in Fig.1. SMC-800-0.83 shows a narrow pore size distribution typical for a 2D hexagonal ordered mesoporous carbon. Lowering the p/s ratio has an effect on the interfacial curvature, shifting 2D hexagonal structures to 3D porous networks with higher pore volumes [35,36]. Decreasing the p/s ratio to 0.5 resulted in a profound broadening of the pore size distribution ranging from 4 to 40 nm. A decrease to 0.25 further increased this effect creating a material with a broad pore size distribution with a pore size range from 4 to 80 nm. This means that the


Page 10 of 29

Page 11 of 29

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

lowest p/s ratio (0.25) was able to create a porous carbon with accessible micro-, meso- and macropores.

Fig. 1. Nitrogen sorption isotherms (left) and PSD (right) of AC and SMC-800-y materials.

The specific surface area of SMC-800-0.83, 0.50 and 0.25 were respectively 670, 663 and 534 m²/g (Table 1). No clear effect of p/s ratio or carbonization temperature was observed on the specific surface area which varies between 422 m²/g and 670 m²/g as seen in Table 1. The specific surface area of AC was 1027 m²/g which is higher than the mesoporous carbons. The mesoporous volume follows the same trend as the pore size and increases by decreasing the p/s ratio as seen in Fig. 1 and Table 1. The pore volume almost doubles from 0.74 cm³/g to 1.37 cm³/g by decreasing the p/s ratio from 0.83 to 0.25. The extra pore volume is created by a change in the mesoporous volume whilst the microporous volume remains almost constant.

Table 1. Material characteristics of AC and SMC-x-y materials.

Material

(m²/g)

(m3/g)

(cm³/g)

(cm³/g)

AC SMC-400-0.83

1027 506

0.50 0.74

0.26 0.69

0.24 0.05

Dp BJH (nm) 2.3 7.0

C (wt%)

H (wt%)

N (wt%)

O (wt%)

86 77

0.3 3.3

4.5 1.5

9.1 18.7


Langmuir

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

SMC-400-0.50 SMC-400-0.25 SMC-800-0.83 SMC-800-0.50 SMC-800-0.25 SMC-1200-0.83 SMC-1200-0.50 SMC-1200-0.25

550 558 670 663 534 631 454 422

1.36 1.56 0.74 1.26 1.37 0.79 1.08 1.31

1.28 1.48 0.60 1.09 1.24 0.67 1.00 1.23

0.08 0.08 0.14 0.17 0.13 0.12 0.08 0.08

Page 12 of 29

29.4 40.0 6.2 23.7 43.8 6.0 21.5 39.9

77 82 91 95 92 92 95 93

3.7 3.9 1.3 0.5 0.5 0.3 0.1 0.3

1.7 0.0 1.2 1.5 1.8 1.1 2.3 2.2

18.1 13.8 7.0 2.8 5.8 6.4 2.4 4.2

The ordered 2D hexagonal structure of SMC-800-0.83 was validated by SEM (Fig. 2). XRD patterns of SMC-1200-y are shown in Fig. 3, the large (001) peak is visible for all three materials indicating mesoporosity. No higher order diffraction peaks are visible due to the broad pore size distributions. The (100) peak of SMC-1200-0.25 appears at a lower angle than for SMC-1200-0.5, and SMC-1200-0.83 has the highest angle. The angle is correlated to the cell parameter and pore size of the material. Higher angles are the result from a smaller pore size.

Fig. 2. SEM pictures of SMC-800-0.83 (left) -0.50 (middle) -0.25 (right).


Page 13 of 29

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

Fig. 3. Powder X-ray diffraction patterns for AC and SMC-1200-y materials.

3.2. Effect of carbonization temperature on material characteristics The effect of carbonization temperature was investigated for the different p/s ratios. At 400 °C a mesoporous polymer material is formed and the transformation to a carbon material is not yet complete. SMC-400-y materials show a distinct brown color instead of the black carbon of SMC-800-y and SMC-1200-y. By examining the elemental composition in Table 1, a difference in mass percentage C is observed between the mesoporous polymers and carbons. The mass percentage C increases with carbonization temperature. However the difference between 400 °C and 800 °C, going from 77 mass percentage C to more than 90 mass percentage C, is more noticeable than the effect from 800 °C to 1200 °C (between 92-95 mass percentage C). This increase in mass percentage C is accompanied with a decrease in mass percentage O. The high temperatures under inert atmosphere remove oxygen containing functional groups such as hydroxyl and carboxyl groups. This increases the hydrophobicity of the material, which in turn is expected to impact on its adsorption capabilities [10,14]. The temperature effect on the structural properties was investigated with nitrogen sorption in Fig. 4. For the SMC-x-0.83 materials, some


Langmuir

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

shrinkage could be observed when transforming from a polymer network to the carbon framework reducing the pore size with almost 1 nm. For the 0.5 ratio and 0.25 ratio, the mesoporous volume also decreased with increasing carbonization temperature. The difference in pore size between 800 °C and 1200 °C is minimal as seen from the pore size distribution in Fig. 4. Shrinkage of mesoporous polymers has been reported to start at temperatures exceeding 400 °C and can range from 1 up to 5 nm depending on the stability of the carbon framework [37–41]. An increased carbonization temperature also induced a mesopore volume loss from respectively 1.48 cm³/g to 1.23 cm³/g (Table 1). In conclusion, for all p/s ratios a carbonization at 800 °C gave good material characteristics such as high pore volume, and broad pore size distribution and a mass percentage C above 90%. Carbonization at a temperature of 1200 °C did not profoundly enhance the material characteristics. Therefore, the indepth HA adsorption study was performed with SMC-800-0.25.


Page 14 of 29

Page 15 of 29

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

Fig. 4. Nitrogen sorption isotherms at 77K (left) and PSD (right) for all SMC-x-y materials.


Langmuir

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

3.3. Humic acid adsorption 3.3.1 Adsorption capacity of all SMC-x-y materials The potential use as an adsorbent for the large humic acid (HA) was investigated for all SMC-x-y materials. The adsorption experiments were performed at neutral pH values, which is typical for water purification plants. The adsorption capacities at initial concentrations of 500 mg/L HA are shown in Table 2. They increased with a decrease in p/s ratio. The carbons with large pores, broad pore size distribution and large pore volume were significantly better at adsorbing HA. This trend was observed regardless of the carbonization temperature, the 0.25 ratio always adsorbed more HA than the other ratios. For a carbonization temperature of 800 °C, adsorption capacities for the ratios 0.83, 0.5 and 0.25 were respectively 260, 294 and 352 mg/g. The effect of carbonization temperature is most pronounced between 400 °C and 800 °C which increased the adsorption capacity from 155 to 260 mg/g for the 0.83 ratio; from 221 to 294 mg/g for the 0.5 ratio; and from 250 to 352 mg/g for the 0.25 ratio. The more hydrophobic carbon is beneficial for the interaction between the adsorbent surface and HA. This supports the idea that most adsorption will occur via π-stacking of the aromatic rings of HA with the carbon surface at neutral pH conditions.

The significant material characteristics were investigated by constructing a model via stepwise regression. The specific surface area, micro and mesoporous volume, pore size and mass percentage C were selected as factors. Only the pore size and mass percentage C were significant factors with p < 0.05. With these factors, a model (Eq.2) was constructed with an R² of 0.95 that is able to predict the experimental adsorption capacities, as shown in Table 2:


Page 16 of 29

Page 17 of 29

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

= −365 + 2.22 + 6.6 !"# $ (2)

Table 2: Experimental and predicted (Eq.2) adsorption capacities (10 mg adsorbent in 50 mL of 500 mg/L HAsolution, RT, 200 rpm, pH = 7), together with the significant material characteristics pore size Dp and wt% C.

Material AC SMC-400-0.83 SMC 400-0.5 SMC-400-0.25 SMC-800-0.83 SMC-800-0.5 SMC-800-0.25 SMC-1200-0.83 SMC-1200-0.5 SMC-1200-0.25

Dp (nm)

Mass percentage C (%)

Exp. qe (mg/g)

Predicted qe (mg/g)

2.3 7.0 29.4 40.0 6.2 23.7 43.8 6.0 21.5 39.9

86 77 77 82 90 95 92 92 95 93

189 ± 17 155 ± 2 221 ± 20 261 ± 1 260 ± 11 294 ± 17 352 ± 20 275 ± 1 312 ± 13 332 ± 1

209 160 209 266 244 316 341 257 305 339

The adsorption capacity of HA on these mesoporous carbons can be predicted via Eq.2 in which only the average pore size and mass percentage C contribute to an increased adsorption capacity. An increase in pore size or mass percentage C will have a beneficial effect on the adsorption capacity. The model clearly indicates the importance of both large pore size and high carbon mass percentage of the carbons for increased adsorption of HA. Larger pore size could allow larger molecules of HA to enter the material and be adsorbed, while the high C mass fraction indicates a more hydrophobic surface, which could improve the interaction with the hydrophobic parts of the HA molecules.

3.3.2 Adsorption kinetics of SMC-800-0.25 and AC The experimental HA adsorption kinetics of SCM-800-0.25 and AC are shown in Fig.5. Because the pseudo second order kinetic model has been used frequently to describe HA adsorption kinetics [26,28,29,31,32], this model was fitted on the


Langmuir

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

experimental data (Fig. 5). The pseudo second order fit had an R² of 0.90 for AC and 0.925 for SMC. The higher k2qe value of SMC-800-0.25 versus AC clearly showed a faster HA adsorption on SMC-800-0.25. The presence of large mesopores clearly increases the HA adsorption rate. The Weber Morris model was investigated as well to identify intraparticle diffusion effects [42,43]. As seen in Fig. 5, the Weber Morris plot shows multilinearity. The first linear part is typically related to diffusion into macro- and mesopores, the diffusion coefficient of this part (kd,1) is larger for SMC800-0.25 than for AC [29,31]. When observing the second linear part, typically attributed to diffusion in micropores, the diffusion coefficients (kd,2) of both AC and SMC-800-0.25 are almost equal. Both plots do not intersect at zero, which is an indication that boundary-layer diffusion still plays a role [31]. The HA adsorption kinetics are therefore influenced by three effects: the boundary layer diffusion, the diffusion in meso- and macropores, which is faster for the SMC-800-0.25 with large mesopores compared to AC, and the micropore diffusion which is the same for both materials.

Fig. 5. Adsorption kinetics of 50 mg SMC-800-0.25 and AC in 500 mL 40 mg/L HA-solution at RT, 700rpm and pH of 7. The data are fitted to the pseudo second order model (left) and the Weber-Morris model (right).


Page 18 of 29

Page 19 of 29

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

Table 3: Kinetic parameters for pseudo second order and Weber Morris model for SMC-800-0.25 and AC.

Material

SMC-800-0.25 AC

Pseudo second order k2 qe R² g/(mg min) (mg/g) 0.000227 27.3 0.925 0.000207 20.6 0.904

kd,1 mg/(g min1/2) 0.620 0.312

Weber Morris C1 R² (mg/g) 5.79 0.975 5.63 0.978

kd,2 mg/(g min1/2) 0.917 1.02

3.3.3 Adsorption isotherms of SMC-800-y and AC Both mesoporous carbons with larger pores, SMC-800-0.5 and -0.25, were selected for further HA adsorption tests. Their experimental adsorption isotherms are shown in Fig. 6. The studied AC reaches an adsorption capacity of 200 mg/g, which is quite high compared to other reported AC [31,32,44]. The AC used in this study has an average maximum volume at pore size of 2.3 nm with some pores up to 4 nm, this is already in the small mesoporous range which could allow the adsorption of smaller HA molecules. However, the mesoporous carbons performed much better with an adsorption capacity of 364 mg/g and 420 mg/g for SMC-800-0.5 and -0.25 respectively. Moreover, the conclusion that the pore size affected the HA adsorption, is validated by the adsorption isotherms: as the average pore size increases from AC over SMC-800-0.5 to SMC-800-0.25, the adsorption capacities increase as well. The larger mesopores of SMC-800-0.25 are able to adsorb more HA than the smaller mesopores of SMC-800-0.5. HA adsorption isotherms are often described by the Langmuir and Freundlich equations [28,29,32]. Both Freundlich and Langmuir fit the data quite well (Table 4). The Langmuir fit is slightly better. Given the large size of HA molecules, the adsorption will probably be limited to a monolayer. The maximum adsorption capacities of the fitted Langmuir isotherms are 236, 470 and 540 mg/g for respectively AC, SMC-800-0.5 and -0.25, which indicates the advantage of a larger


C2 (mg/g) 19.8 12.4

R² 0.936 0.975

Langmuir

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 20 of 29

pore size on the adsorption of HA. The adsorption capacity of SMC-800-0.25 exceeds those of reported AC (52.4 mg/g), as well as the hard templated CMK-3 material (137 mg/g) which were tested at pH 7. The adsorption capacity of soil mineral bentonite at 537 mg/g is comparable to SMC-800-0.25, however this value was obtained at a pH between 3 and 4. Higher pH values resulted in a drastic lowering of the adsorption capacity on the bentonite. For SMC-800-0.25 the adsorption capacity at lower pH values (3-4) will be even higher compared to the adsorption capacity at pH of 7 (Fig.7). Table 4: Langmuir and Freundlich parameters for SMC-800-0.25, SMC-800-0.5 and AC.

Material

Langmuir KL

qmax

Freundlich R²

KF 1/n

1/n

R²

1/n

(L/mg)

(mg/g)

SMC-800-0.25

0.00370

540

0.985

(mg L )/(g mg ) 15.9

0.490

0.978

SMC-800-0.5

0.00340

470

0.996

12.2

0.509

0.973

AC

0.00440

236

0.986

9.64

0.451

0.968

Fig. 6. Adsorption isotherms of SMC-800-0.25, SMC-800-0.5 and AC with Langmuir fit, 10 mg of adsorbent was added to 50 mL HA-solution of various concentrations at 25 °C, 200 rpm for 7 days.


Page 21 of 29

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

3.3.4 Influence of the pH The mass percentage C of the mesoporous carbons significantly affected the adsorption capacity, as shown in Eq. 2. This was explained by the hydrophobic interactions between HA and the adsorbent being the main adsorption interaction. To confirm the hydrophobic adsorption interaction, the effect of pH on the HA adsorption on SMC-800-0.25 was studied at different initial pH values (3, 5, 7, 9 and 11). As seen in Fig. 7, the HA adsorption increases dramatically at lower pH values. At low pH, the electrostatic repulsions between the HA and the carbon surface decrease because the carboxyl and phenolic hydroxyl groups of HA are protonated [45] and the negative charge on the carbon surface becomes smaller (Fig. S1). This allows more hydrophobic interactions and, hence, increases the adsorption [24]. Upon a pH increase, the carboxyl groups deprotonate first (pKa~5) and at higher pH values the phenolic hydroxyl groups (pKa~10) deprotonate as well. This makes HA more hydrophilic, causes electrostatic repulsion with the hydrophobic and more negatively charged carbon surface and, hence, decreases the adsorption [28,43,46]. To achieve the highest adsorption capacity with these mesoporous carbons, working at low pH (

Tunable Large Pore Mesoporous Carbons for the ... - ACS Publications

Recommend Documents